Hi,
I wondering if anyone has insight into Multiple Domain SSM optimization parameters for a femur/pelvis/sacrum. I’ve been working for the past few days making changes to the parameters and have not found any that work. I’ve attached a picture of the parameters I currently have. Any suggestions are welcomed!
Thanks!
The femur is usually not too challenging. The pelvis we’ve found benefits from use the of Normals. The sacrum is more challenging and we’ve had luck using the Geodesic Distance option. Be warned that it is significantly slower.
The Relative Weighting of 20 seems a bit high. I would recommend lowering it closer to 1-5.
I’ve been able to get the individual groupings of each bone to work, however they didn’t have the same optimization parameters to plug in to M.D.SSM together. Is there a way for me to individually specify parameters for the bones separately?
For geodesic distance - are there any other parameters that will help make this function work (ie. lower/higher iterations, etc.)? Every time I try to use geodesic distance, I get an error message that won’t let the optimization run.
There is not a way to specify different optimization parameters for different domains. This is partly due to the fact that they run as a joint model and many parameters apply to that joint space.
You can, however, run each domain as a separate optimization and combine the results into a single shapeworks projects spreadsheet. The joint statistics will then be used on that multi-domain model. Just make sure to use the same combination that results from the multiple-domain execution.
What error are you receiving when using geodesic distance?
I figured out the optimization parameters, thank you!
For the geodesic distance - I don’t have a picture of the error message right now, but I will share if I come across it again.
I tested out the geodesic distance function again, as I am working on trying to optimize a data set of structures for multiple domain SSM (I’m working on a set of 11 muscles, trying to find optimization parameters that work for all of them).
The error message that popped up when trying to use geodesic distance was:
Is there a step I might be missing in order to use geodesic? Or is there certain optimization parameters that need to be on/off in order for it to work?
Thanks!
My only guess off of the top of my head is that one of the meshes has some kind of topology problem. There may be non-manifold vertices/edges and the geometry central geodesic computation can’t process it. We are working on some tools that can detect and fixes these types of issues.
If you can, I would try to bisect the data to determine if there are one or more meshes that are causing this.
